Creating your plugin

Plugins are pieces of custom code that you can bring onto the platform. Plugins can be made with many of the popular languages and libraries for data analysis, such as Python. Plugins can be made available publicly to others, or keep them private within your organization's template.

Before your create a plugin, you need to decide on one of the two plugins types:

  1. Report plugin — uses a report as data source. Can be a timeseries (e.g. for forecasting), list of all users (e.g. for a/b testing) or a list of all events (e.g. for anomaly detection). Can be run by end-user from the Plugins drop-down menu on any report results page, or used in the Insight API to glue them together with other components.

    If this is what you need, go directly to the Report plugin documentation.

  2. Intelligence plugin — uses an intelligence model as data source. An intelligence model can have a binary goal (true or false), or no goal for unsupervised models like clustering. A few examples of available intelligence goals are:

  • User did event X.
  • Sum of some event property for user is at least X.
  • User came back after X days.
  • User had X sessions.
  • User complete step X of a funnel.
  • Any custom goal defined by SQL.

    If this is what you need, go directly to the Intelligence plugin documentation.